NHH_Il
Norges -
Haudclshøyskole
Nor weg i anS cho ol of lic onom ic s and Business Administration
This dissertation is the result of a research interest that started in the late 1980s while I was at the Norwegian Fund for Market and Distribution Research. At the time, I wrote a theoretical paper on a ·transaction cost approach to the internationalization of firms for presentation at the National Conference on Business Economics (FIBE-VI) held in Bergen in January 1989. While that particular paper was not taken further it awakened my interest in doing empirical work in the area, and the task of compiling a data base on Norwegian foreign direct investment started later that year. That data base was eventually used in an article written together with Geir Gripsrud at the Norwegian School of Management, which - after a series of revisions - was accepted for publication in
Journal of International Business Studies.
lowe special thanks to Geir Gripsrud. Inaddition to being the start of a productive and enduring research partnership with him, that article became a vital stepping stone toward the completion ofthis dissertation. A special debt of gratitude is also owed to Lawrence S. Welch (now at the Norwegian School of Management). His intellectual inspiration and cooperation, friendship and encouragement has been of great importance throughout the research process.
Sincere thanks goes to Donald Storrie (now at the University of Gothenburg) for all support and encouragement in a critical phase of the work and forhisconstructive comments on parts of this dissertation, to Arne Nygaard (Norwegian School of Management, School of Marketing) for having spent much time discussing my ideas and for providing critical comments on practically every aspect of the research, and to Tom Stranger-Johannessen for his assistance with the data bases, for his many useful suggestions, and for helping me out with the intricacies of preparing the final manuscript. My gratitude also goes to my principal adviser, Kjell Grønhaug at the Norwegian School of Economics and Business Administration whose comments and support at various stages of this work have been greatly appreciated, and to Torger Reve at the Norwegian School of Economics and Business Administration, for his useful suggestions in the later stages of the work.
ii
Colleagues at my fonner workplace at Østfold College, in particular Øystein Strøm, Theo Schewe and Egil Skorstad were always willing to engaging in stimulating discussions. From my present workplace at the Norwegian School of Management, I received valuable inputs from Erik Olson in the later phases of the process, while my other colleagues at the Department of Marketing and Logistics have provided a stimulating work environment. In addition, they ensured that I could spend most of the fall tenn of 1994 concentrating on completion of this dissertation. Peggy Brønn helped improve the English in one of the articles, and the personnel at the Libraries at Østfold College and at the Norwegian School of Management provided excellent service. I thank them all. Finally, a grant from Høgskolestyret i Østfold gave me the opportunity to devote myself exclusively to doing research in the fall tenn of 1993. Their support is gratefully acknowledged.
Oslo, April 1995
Gabriel Robertstad Garcia Benito
Table of Contents
Preface i
List of Tables v
List of Figures vii
Chapterl
Introduction 1
Chapter2
The Expansion of Foreign Direct Investments: Discrete Rational Location Choices or
a Cultural Learning Process? 41
Chapter 3
The Internationalization Process Approach to the Location of Foreign Direct
Investments: AnEmpirical Analysis 67
Chapter4
Ownership Structures of Norwegian Foreign Subsidiaries in Manufacturing ... 97
Chapter 5
Foreign Market Servicing: Beyond Choice of Entry Mode 139
Chapter6
Divestment of Foreign Production Operations 171
Chapter7
Summary and Conclusions 209
iv
List of Tables
Table 1.1. Number and geographical distribution of foreign manufacturing subsidiaries owned by Norwegian companies in 1969, 1974 and 1982 (percentages in
parentheses) '.' 6
Table 1.2. Total number and geographical distribution of foreign manufacturing subsidiaries
owned by Norwegian companies 1984-1987 7
Table 1.3. Norwegian FDI in 1988 to 1992. Figures in bill. current NOK 8 Table 1.4. Number of Norwegian foreign manufacturing subsidiaries in 1984 by
industry 9
Table 1.5. Ownership structures of Norwegian FDIs in manufacturing in 1984 10 Table 1.6. Entry and exit of Norwegian FDI in manufacturing; 1969 to 1992 11 Table 2.1. Geographical distribution of Norwegian FDIs in manufacturing 54 Table 2.2. Average distance of FDIs by sequence of investments 56 Table 3.1. Means, standard deviations, and Pearson correlation coefficients (n
=
203) ., 82Table 3.2. Multiple regression of equations (1)-(3). OL5-estimation (n
=
203) 83 Table 3.3. Multiple regression of equations (1)-(3) for subsample consisting of FDIs withinvestment number five or less. OL5-estimation (n =147) 85 Table 4.1. Summary of hypotheses on choice of wholly-owned subsidiaries 116 Table 4.2. Descriptive statistics (total sample, n
=
174) , 117 Table 4.3. Correlations among independent variables (total sample, n =174) 118 Table 4.4. Results of logistic regression: Maximum likelihood estimation wholly-ownedsubsidiary versus joint venture (total sample, n
=
174) 120vi
Table 4.5. Results of logistic regression: Maximum likelihood estimation wholly-owned subsidiary versus joint venture (horizontal integration sample, n
=
125) . . . .. 122 Table 4.6. Results of logistic regression: Maximum likelihood estimation wholly-ownedsubsidiary versus joint venture (vertical integration sample, n
=
49) 124 Table 6.1. Status in 1992 of the stock of Norwegian FD! in 1982 188 Table 6.2.Descriptive statistics for independent variables (n=
152) 192 Table 6.3. Results of logistic regression: Maximum likelihood estimation of subsidiaries still in operation versus divested subsidiaries (n =152) 194Ta~le 7.1. Summary of empirical findings 215
List of Figures
Figure 1.1. Framework for the studies 24
Figure 2.1. Norwegian FDls in manufacturing by year of investment 53 Figure 5.1. International experience and foreign market commitments: A hypothetical
example. . . .. 147 Figure 5.2. Process and situational influences on internationalization . . . .. 150
Figure 5.3. Foreign market service packages 154
viii
Chapter 1
Introduction
2
Background
The immense growth of foreign direct investment (FD!) in the Post-
wwn
period - in particular during the three last decades - has led to a considerable interest in research about various aspects of FD! and the behavior of multinational enterprises (MNEs). Foreign direct investment entails a cross-border investment made by a company for the purpose of .acquiring a sizeable long-term equity interest in a foreign enterprise, and thereby exert a considerable degree of influence on the operations of that enterprise. Control ~ver operations undertaken in a foreign location isa key feature of FD! (Caves and Jones, 1977). Although the operational definitions of "control" may vary (Cohen, 1975), having 10 per cent or more interest in a foreign business venture is by many - including, for example, the Bank of Norway - regarded as the critical cut-off point. Another key feature of FD! is that it represents, in principle, a long-term commitment to a foreign operation. Foreign direct investment differs from international portfolio investments, i.e. short-term capital movement decisions, from a control perspective as well as from a time-frame perspective. A third characteristic of foreign direct investment is that this type of investment involves the collective transfer of various resources, including factor inputs such as technology, entrepreneurship, and managerial knowledge (Hood and Young, 1979). As noted by Balasubramanyam (1985, p. 161) "the essence of FD! is that it is a package of capital, technology and managerial skills". Thus, FD! serves as an important vehicle for transfers of not only capital, but also of technology and managerial resources between countries. A MNE, which in accordance with this concept of FD! can be defined as a company which owns, controls and manages income-generating assets in more than one country (Hood and Young, 1979), must therefore be regarded as a key actor in any analysis of international economic relations.Although a vast literature exists, our knowledge is still only rudimentary regarding many aspects of FD!. First, while many studies have addressed the question of why companies chose to operate in foreign markets by undertaking foreign direct investments, relatively few
4
studies have looked into how these investments are actually made, for example whether they are made in the form of wholly-owned subsidiaries or as joint ventures. This issue is of particular interest from the viewpoint of the control aspect of FDI; if FDls are undertaken in order to gain controlover foreign operations, why are foreign investors seemingly often willing to share decision authority with an outside partner? Second, the bulk of the literature, especially the literature rooted in economies, takes a rather static view of foreign direct investment behavior'. For example, little is known about how individual companies expand internationally: which routes - in terms of operation modes, locations, etc. - do companies take as they grow into large MNEs? Regarding the dynamics of FDI, an intriguing question is also why and to what extent foreign investments are terminated. FDI represents, as already noted, a long-term commitment to a foreign market. It is, nevertheless, far from unusual (see for example, Boddewyn, 1979) that companies decide to dismantle their engagement in given foreign operations. Still, the issue of foreign divestment is largely unexplored in the literature.
Third, most empirical studies to date have focused on FDI in a North-American context, i.e.
either the foreign direct investment behavior of U.S. companies, or the inflows to the U.S. of FDI made by non-U.S. companies. However, given the special characteristics of the North- American market such as its size, and the substantial financial, managerial and technological resources available to many U.S. multinationals, it is far from clear that the insights gained from studies conducted in a U.S. context can be readily transferred to other contexts.
In this dissertation four empirical studies of the behavior of Norwegian MNEs are presented.
In addition, a conceptual study is included in the dissertation. Throughout the dissertation the focus is on FDI in manufacturing. Such operations are of particular interest due to the substantial commitment of resources involved in setting up or acquiring foreign manufacturing subsidiarie. The studies address issues that have been neglected in previous studies, namely: what impact does company characteristics have on the expansion pattern of foreign direct investments? What factors influence the choiee of ownership structure of foreign subsidiaries, and more generally, how do companies enter and develop operations in foreign markets? Finally, what determines companies' exit from given foreign operations?
This dissertation explores these issues from the perspective of two central streams of
literature on the internationaloperations of firms: the theory of the multinational enterprise (see for example Dunning, 1993) and the theory of the internationalization of the firm (Welch and Luostarinen, 1988). However, before delineating the theoretical frameworks and research problems investigated here in greater detail, a brief account of Norwegian FD! in manufacturing will be given as an introduction the empirical context to be studied. A brief presentation of the studies closes this chapter.
An Overview of Norwegian Foreign Direct Investment
Norwegian foreign direct investment has traditionally been quite modest. The first Norwegian company to establish production plants in foreign countries was O. Mustad &
Søn A/S, which in the early years of this century began production of fishing equipment "in several other European countries; Sweden, Germany, Great Britain, France, Italy, Spain and Austria-Hungary (Hodne, 1993). O. Mustad & Søn A/S and a few other Norwegian companies, such as A/S Borregaard, O. Kavli A/S and Wessel & Co. A/S, were however exceptions to the overall picture of a largely domestic oriented Norwegian manufacturing industry until the 1960s.
In the sixties, two important events led eventually to major changes in Norwegian companies' outward orientation. First, an era of trade liberalization commenced with the EFTA agreement signed in 1960, which gradually opened-up the previously protected Norwegian market. The removal of tariffbarriers for manufactured goods exposed indigenous companies to competition from imported goods, which in turn forced many companies out of business (Hodne, 1993). Moreover, those that were able to stay in business, usually by focusing on relatively narrow product niches, were hampered by the small size of the domestic market.
Thus, expansion into foreign - typically neighboring - markets became the key to survival.
Second, the discovery and subsequent development of significant oil reservoirs in the North Sea, albeit beneficial to certain industrial sectors (especially those that were related to the exploration and operation of oil fields), was a major contributing factor to the escalation of
6
costs in Norway. Consequently, an outflow of FDI, which to a great extent was based on cost considerations, took place from the late sixties and throughout the seventies. Southern Europe, Southeast Asia, and the Middle East, were target areas for these investments (Carlsen and Rasmussen, 1988; Hodne, 1993, Smukkestad, 1979).
Following these developments a somewhat restrictive national policy on FDI was gradually replaced by a more liberal view on capital movements (NOU, 1981; St.prp. (1984-85);
Johansen and Holm, 1989). In 1984, restrictions on inward FDI were formally lifted, and some years later restrictions on outward FDI followed suit. The growth of Norwegian FDI reflects by and large these developments regarding trade regimes, industrial structural change, and national policy. In 1969, an overview compiled by the Norwegian Industrial Federation - the only fairly comprehensive FDI statistics available at the time - identified 86 foreign.
manufacturing subsidiaries owned by Norwegian companies (see table 1.1.). The number of subsidiaries had increased to 131 in 1974, and to more than two hundred in 1982.
Table 1.1. Number and geographical distribution of foreign manufacturing subsidiaries owned by Norwegian companies in 1969,1974 and 1982
(percentages in parentheses).
~
Region 1969 1974 1982
Scandinavia 29 (34%) 47 (36%) 66 (33%)
Europe 27 (31%) 41 (31%) 59 (29%)
North-America 9 (10%) 9 (7%) 35 (17%)
Other 21 (24%) 34 (26%) 41 (20%)
Total 86 (100%) 131 (100%) 201 (100%)
(Sources: Norges Industri (1969),Norges Industri (1974),Norges Industri (1982))
The period from the mid-eighties onwards has witnessed, in general, a move toward an even more liberal trade and investment climate worldwide, but has also been characterized by increasing regional economic integration, e.g. EU, NAFTA (Dunning, 1992). In addition - or perhaps, as a result - in many industries the global competition among oligopolistic MNEs
induced a wave of international mergers and acquisitions whereby companies attempted to attain the necessary scale and scope economies to retain their market presence (Dunning, 1993). As a result, FDI activity increased considerably throughout the period. Norwegian companies, of which several now can be considered as true multinationals (Hodne, 1993), took part in this devel~pment. First, in the 1980s, the number of foreign manufacturing subsidiaries of Norwegian companies rose quickly, reaching 547 in 1985 (Norges Bank, 1986) and 681 in 1987 (Eksport-aktuelt, 1988). Moreover, the regional distribution of FDI changed somewhat compared to previous periods. As seen from table 1.2., a steadily increasing proportion of Norwegian foreign establishments were made in Europe. Thus, it may seem that the economic integration developments that took place in Europe from the mid-1980s onwards attracted many Norwegian companies to establish or reinforce their presence in that . area.
Table 1.2. Total number and geographical distribution of foreign manufacturing subsidiaries owned by Norwegian companies 1984-1987.
~
1984 1985 1986 1987
In total 385 (100%) 547 (100%) 656 (100%) 681 (100%)
ofwhich:
Scandinavia 42% 38% 39% 38%
Europe 31% 35% 36% 37%
North-America 10% 10% 10% 9%
Other 17% 17% 15% 16%
(Sources: Hansen (1984), Norges Bank (1986, 1987),Eksport-aktuelt (1988».
Conversely, while the number ofFDls undertaken in countries outside Europe and the Ll.S.
certainly increased from 34in1974 to 66 in 1984, and 107 in 1987, the relative number of FDls made in those countries showed a downward trend. This suggests that - for Norwegian companies - cost considerations may, in general, have been less important as a determining factor in the location of FDI in the 1980s than they apparently had been just a decade earlier.
8
There are no readily available surveys of the number of production units owned by Norwegian companies in recent years. However, FDI statistics have been compiled by Norges Bank based on the amount of long-term capital invested in foreign countries. Table 1.3. shows the stock of total Norwegian FDI and FDI in manufacturing in the years 1988 to 1992. The figures reveal that Norwegian FDI continued to increase between 1988 and 1992.
Table 1.3. Norwegian FDI in 1988 to 1992. Figures in bill. current NOK.
Year FDI in manufacturing Total FDI
1988 11.9 26.1
1989 15.8 33.7
1990 19.2 40.5
1991 20.9 46.0
1992 24.4 47.7
(Source: Norges Bank, 1993)
To summarize; from a modest start in the mid-sixties, Norwegian foreign direct investment has increased impressively during the last 25 years. There are, nevertheless, still only a limited number of large companies by international standards. For example, the total number of Norwegian companies owning at least one foreign manufacturing subsidiary in 1984 was only about one hundred (Norges Industri, 1984). Hodne (1993) - using a more strict definition of a multinational enterprise (at least three foreign production plants, and sales of a minimum of two billion NOK) - finds that only eleven Norwegian companies belonged to the group of
"true" MNEs in 1991.
This concentration of FDI is also reflected in the sectoral distribution of Norwegian foreign direct investment in manufacturing. As shown in table 1.4. more than half of the foreign subsidiaries existing in 1984 belonged to only two main industrial sectors; chemicals and machinery.
Table 1.4. Number of Norwegian foreign manufacturing subsidiaries in 1984 by industry"),
Industry (two-digit SIC) Frequency Percent
Food, beverages and tobacco 20 7.9%
Textiles 14 5.5%
Wood products and furniture 4 1.6%
Paper products 14 5.5%
Chemicals 60 23.6%
Mineral products 12 4.7%
Basic metals 22 8.7%
Fabricated metal products and machinery 89 35.0%
Other manufacturing industries 4 1.6%
Notknown 15 5.9%
Total 254 100.0%
*)Based on data fromNorges Industri (1984).
The remaining FDIs were more evenly distributed across industries. Looking at the geographical distribution of Norwegian FDI it is evident that besides some interest in low- cost production sites, particularly in the seventies, other Scandinavian and European countries have been the main targets for Norwegian foreign investment. There are probably many reasons for this; geographical and cultural proximity, as well as the fact that demand exists for relatively highly priced products are important. Furthermore, trade links were in many cases already established. The fact that Norway was not, and still is not, a member of the EC (now the EU) may also contribute to the investment pattern. Given the importance of the European markets to many Norwegian companies, the economic integration that has taken place in Europe in the last decade probably has provided a range of strategic (for example achieving sufficient scale economies) as well as market access (for example trade impediments such as local content and local production requirements) motives to undertake FDI in that area.
What has been described so far is the overall picture regarding Norwegian foreign direct investment. Development processes at the level of individual investor companies and the
10
various foreign subsidiaries - such as routes of expansion regarding the location of FDIs - and the characteristics of these companies have not been covered. However, such topics will be discussed in more detail in the studies included in this dissertation, and this short overview of Norwegian FDI is therefore limited to cover only some of these aspects. An important dimension of FDI is the ownership arrangement of the subsidiaries. Although FDIs are per definition oriented toward controlover operations undertaken in a foreign location, there is considerable latitude for how much control an investor actually has in a given foreign operation. A crucial factor is the percentage of equity owned by a focal company, which can range from 10percent to 100percent.
Table 1.5. Ownership structures of Norwegian FDIs in manufacturing in 1984*).
Equity percentage Frequency Percent
10-49 (minority) 50-50 (balanced) 51-99 (majority) 100 (wholly-owned) Notknown
64 12 38 128 12
25.2%
4.7%
15.0%
50.4%
4.7%
Total 254 100.0%
*)Based on data from Norges Industri (1984).
Table 1.5.shows the distribution across various categories of ownership for Norwegian FDI existing in 1984. Most of the FDIs fall into two groups; about half of the FDIs are wholly- owned, and about one quarter belong to the minority-holding category. A comparison with U.S. data suggests that there are noteworthy differences between Norwegian and U.S.
companies. A study done by Gatignon and Anderson (1988) indicated that wholly-owned arrangements are used more frequently by U.S. companies: wholly-owned subsidiaries accounted for more than 70 percent of the number of FDIs in manufacturing made by U.S.
companies. Thus, while there is a certain preference for wholly-owned subsidiaries among Norwegian companies, it is not as clear as for U.S. companies. Many companies have
apparently not wished to, or been able to, obtain complete ownership controlover their foreign subsidiaries. This dissertation will examine why.
Another interesting issue pertains to the question of entry and exit dynamics which, partly because of difficulty in getting data, has largely been overlooked in previous studies. Table 1.6. presents some information on the number of new subsidiaries and the dissolution of existing subsidiaries for various periods from 1969 to 1992. The table is based on data from the surveys undertaken by the Norwegian Industrial Federation (1969,1974 and 1982) and data collected through inspection of annual reports of the companies (1992f Although the data do not provide the complete picture, the table clearly indicates that dissolutions, or divestments, have been a rather common phenomenon throughout the period. For example, almost 27 per cent of the foreign manufacturing subsidiaries owned by Norwegian companies in 1974 were divested over a period of only eight years (calculated as [01982 / A1974] x lOO, table 1.6.). The exit rate for the following ten-year period was even more dramatic: more than half (53.2 percent) of the manufacturing subsidiaries owned by Norwegian companies in 1982 were divested in the following ten-year period (i.e. [01992 / A1982] x 100). Moreover, it should be noted that the calculations presented here are - due to lack of data on entries and exits in the in-between periods - in fact quite conservative estimates of the actual number of divestments.
Table 1.6. Entry and exit of Norwegian FOI in manufacturing; 1969 to 1992 ').
Xfgr
1969 (tI) 1974(t2) 1982(t3) 1992(t4) U)
A.in existence
t
n 86 131 201B.remaining from
t
n_1 70 96 74C. new during
t
n -t,
-1 61 105O. divested during tn - tn-1 16 35 107
') Based on data fromNorges Industri (1969, 1974, 1982) and companies' annual reports (1992).
") Data missing on 20 subsidiaries belonging to the 1982-1992 data set
12
Theories of the Multinational Enterprise and the Internationalization of Enterprises
Two theoretical perspectives - the "eclectic" theory of international production and the
"internationalization process" model - dominate at present in the analysis of FDI and internationalization in general. The so-called "eclectic" theory of international production developed by Dunning (1980,1988,1993), draws on industrial organization and transaction cost theory together with elements of location theory in order to provide a general approach to foreign direct investment behavior. This approach places particular importance on the various market imperfections that companies face while undertaking different types of cross- border transactions, and contends that these market imperfections are by-passed by relying on internal instead of external cross-border transfers of input factors, semi-finished goods, and/ or finished goods. Inorder to create an internal channel for such transfers a foreign direct investment must be made.
The behavioral theory of the firm provides the basic conceptual elements underlying the so- called "internationalization process" model developed primarily by Nordic scholars like Luostarinen (1970, 1979), Carlson (1975), and Johanson and Vahlne (1977). The "process"
approach looks at the internationalization of companies, including their foreign direct investment behavior, through the lens of organizational decision-making behavior, in which concepts like uncertainty, limited search, and experientiallearning are central. The main prediction of the model is that the various aspects of the internationalization of firms take place along a path of gradual development.
Several comprehensive surveys of the theories of foreign direct investment and the multinational enterprise have been undertaken recently (see for example Cantwell, 1991).
Also, both Welch and Luostarinen (1988) and Johanson and Vahlne (1990) provide useful overviews of the main work done in the area of the internationalization of the firm. The short
review undertaken here will therefore be confined to outlining the main concepts and ways of reasoning embedded in these two streams of literature.
The "eclectic" theory of international production outlines a generalized framework for explaining the level and pattern of the cross-border activities of firms (Dunning, 1993, p. 79).
His framework brings together previous work on foreign direct investment and the multinational enterprise dating back to Hymer (1960). In his seminal work on the multinational enterprise, Hymer observed that indigenous firms, at least initially, have advantages over foreign firms in the domestic market, because of their superior knowledge of the home country and the home market, and because these firms already have undertaken the investments (such as setting up a distribution system) needed for serving the market. In . order to compete with domestic firms, foreign companies must therefore have some 'advantage that compensate for the disadvantages (which can be conceptualized as entry barriers) they face when operating in a foreign environment. A number of advantages, which by nature are of a monopolistic kind, were identified by Hymer and others. Important are superior technology and managerial skills, cheaper access to capital, economies of scale (Hymer, 1960; Kindleberger, 1969), differentiated products and brand names (Caves, 1971), and technological know-how and skills (e.g. Johnson, 1975).
However, possession of monopolistic advantages does not provide a sufficient condition for FDI to arise since one may ask why a firm having such advantages does not attempt to have
"the best of both worlds", that is, transferring their technology while avoiding the costs of doing business abroad, by licensing their technology to indigenous firms (Casson, 1987).An answer to this question was worked out by a number of writers, especially MacManus (1972), Buckley and Casson (1976), Magee (1977), Rugman (1981), Williamson (1981), and Hennart (1982), building on the insights regarding the nature of the firm originally put forward by Coase (1937). Although the terminology differs somewhat, the basic point made by all contributors is that transfers of the assets in question - which often are of an intangible nature - are prone to encounter severe problems due to various types of market failure. First, the terms of a contract may be difficult to determine ex ante if the asset to be transferred is
14
complex, and/ oriftransfers require a close co-ordination of the activities of buyers and seller over a lengthy period of time (for example, the development of a specially designed intermediate product or component). This problem is basicallyone of uncertainty; it is difficult to foresee future outcomes (it might even be difficult to define accurately the present state-of-nature). Second, even ifthe terms of a contract were agreed upon, the parties to the contract still are exposed to the hazards that breaches of contract might not be fully compensated. This problem isbasically one of opportunism
ex post
(Williamson, 1981), which can be of great importance in cross-border transactions since judicial enforcement often isless effective in an international context. Third, the assets constituting the monopolistic advantages of the firm (e.g. technological know-how) can be difficult to transact due to the 'buyer uncertainty" problem, i.e. that the seller knows the value of the asset, while the buyer cannot appraise itex ante.
Clearly, a complete disclosure (for example by providing the asset ..on trial) could of course mitigate that problem. However, since such assets often have an
"information good" character, i.e. once disclosed the buyer would no longer have an incentive to pay for the asset since he already has acquired it, on-trial evaluations are not satisfactory.
The patent system may of course solve the problem, but to what extent patents are enforced varies considerably across countries. Finally, assets may be of an intangible nature (e.g. tacit knowledge) which makes them difficult to codify (and hence protect by, for example, a patent), and difficult to transfer in an immediate, once-for-all fashion. Whenever any of these circumstances occur the costs of operating in external markets for transactions of these kinds are bound to outweigh the benefits (or external markets may even not exist due to the high costs). Instead, companies develop their own internal organizational structures to achieve internal co-ordination of activities. From this perspective the MNE then becomes a special case of the more general "boundaries of the firm" problem, or to quote Buckley and Casson (1976, p. 45): "a MNE is created whenever markets are internalized across national boundaries".
Still, the question of why a firm would undertake production abroad instead of producing for export from the home country remains unanswered. In other words, what explains the location pattern of international production? According to economic theory, the location of
production is mainly determined by comparative advantages, as in trade theory, and by barriers to trade (Clegg, 1992). The theory predicts that companies would choose locations that minimize the total costs of manufacturing the product and delivering it to the buyer. An important location advantage associated with a particular country is the availability of inputs, such as natural resources, to the firms established in that country. Inaddition, the costs of producing in a specific location are influenced by numerous other factors including labor cost differentials, transportation costs, realization of scale economies (which can be a function of market size, and consequently of market growth), government policies regarding taxes (or conversely; subsidies), and trade barriers implemented by the governments such as quotas, tariffs and "local content" requirements. These factors influence location costs in a fairly straightforward and - at least in principle - quantifiable way (especially in a static analysis) ..
Finally, the attractiveness of various locations may also hinge on factors that while they have an impact on costs, that may be far more difficult to calculate. Inparticular, the management of foreign operations is
per se
likely to incur costs (Hirsch, 1976), because of, for example, communication difficulties. Such costs are likely to be less in familiar markets, that is, markets that culturally and socially resemble the home market or markets with which the company has previous experience.Pulling these streams of research together Dunning (1993) then suggests that the three following factors must be taken into consideration in order to explain why companies are capable of, and would chose to, operate value-added activities (e.g. sourcing, manufacturing, marketing) in foreign countries:
i) A company possesses net ownership advantages versus firms of other nationalities in serving particular markets. These ownership advantages are firm-specific in the sense that the firm, at least for a period of time, has controlover them. They include patents, know-how and possession of superior production technology, controlover markets, scale advantages, managerial capabilities, specialized labor skills, etc. These factors determine together the competitive position of a firm in relation to other firms.
16
ii) Given that ownership advantages are present, it must be more beneficial to the company possessing these advantages to use them itself rather than to lease them to foreign firms to use, i.e. to internalize the advantages through an extension of its own activities rather than externalize them through arm's-length transactions with some other independent firm. Internalization advantages arise from the existence of various market imperfections as analyzed above.
iii) Finally, given internalization of ownership advantages, it must be beneficial for the company to use these advantages in conjunction with at least some factor inputs (e.g.
natural resources) outside its home country. Otherwise foreign markets would be served entirely by exports, and conversely, the domestic market would be served by local production.
As noted by Buckley (1988), the theory rests implicitly on the assumption that firms rationally take into account the factors included in the theory, i.e. that they in any given situation correctly assess their competitive position relative to other firms, that they can calculate the comparative costs of different governance structures for international resource transfers, and that they chose the least-cost locations for undertaking given value-added activities, and furthermore, that all these considerations are made in conjunction. However, while the
"eclectic" theory of international production may provide an explanation - on the basis of rational behavior - to why foreign direct investment takes place, and thereby why MNEs exist, it has little to tell about how companies have achieved a position in which the outlined conditions for FDI apply. Hence, even if one accepts the perhaps overly optimistic view on the degree of rationality implied by the economics approach to FDI, a number of questions remain unanswered. For example, how do firms acquire the advantages necessary to compete in foreign markets, and in which circumstances are firms lead to a situation where they face the decision to make a FDI or not in the first place? Inaddition, the rejection of the notion of rational decision making, which has been challenged by numerous scholars of business behavior (see for example March, 1978), provides an alternative platform for studies of the international behavior of firms.
Afirst step towards an understanding of the organizational processes behind the decision to make a FDI and where to make the FDI, was undertaken by Aharoni (1966) in his study of the various steps involved in the foreign direct investment decision process. Aharoni demonstrated that the decision making process resulting in FDI had little resemblance to rational decision making. First, rather than being the outcome of a continuous quest for an optimal fit between the resources and competencies of the company and the opportunities provided by the environment, the process often was initiated by external forces (such as investment proposals presented to the companies by foreign governments, the distributors of the companies' products, or even their customers), the conviction held by some executive that FDI should be made, or was triggered as purely imitative behavior, for example by observing that a competitor had successfully served·a market by FD!. Furthermore, Aharoni's study suggested that once a FDI opportunity was identified, the subsequent decision making process involved limited, if at all, evaluation of alternatives. Instead, the companies tended to treat investment opportunities on a singular and dichotomous (invest / do not invest) fashion. Whether decisions were reached as to go-ahead with an investment was dependent on sufficient organizational support for that specific course of action, which in tum was largely determined by who advocated the idea internallyand when the proposal was being considered.
Similar notions of constrained decision-making behavior were suggested by findings in a number of studies looking at the export behavior of firms. The decision to start exporting was rarely planned, but rather externally initiated, for example by unsolicited orders (Wiedersheim-Paul et al.,1978; Bilkey and Tesar, 1977) and government export stimulation activities (Wiedersheim-Paul et al., 1978), or triggered by situational internal factors such as surplus production (Tookey, 1969) and the existence of entrepreneurial individuals within the firm (Simmonds and Smith, 1968). Moreover, Lee and Brasch (1978) noted that in many cases such decisions involved little collection of information, and were made without a clear definition of goals. Finally, a number of studies suggested that the internationalization of the firm could be described as a process consisting of a series of small steps, whereby firms gradually increased their international involvement (Welch and Luostarinen, 1988; Bilkey and
18
Tesar, 1977; Johanson and Vahlne, 1977;Johanson and Wiedersheim-Paul, 1975), rather than full-fledged internationalization from the outset. Several models of how the internationalization of firms proceeds have been presented in the literature, but as pointed out by Andersen (1993) the similarities between the various models outnumber the real differences between them. Taken together these studies established what might loosely be called the "internationalization process framework".
The internationalization process framework is rooted in a behavioral decision making approach. This perspective on the internationalization of the firm applies the concepts of bounded rationality (Simon, 1955), uncertainty avoidance, limited search (March and Simon, 1958), and organizationalleaming (Cyert and March, 1963) in the analysis offirms'behavior.
The stages in the decision process are discussed in detail by Luostarinen (1979 who, building.
upon Cyert and March (1963), underlines the importance of "lateral rigidity" between the stages in the decision process: limited perception of alternatives and selective search leads to confined choice. Hence, in the initial stages firms are predicted to enter nearby markets by means of low-commitment operation methods such as exporting to a local representative.
However, as more knowledge is acquired more alternatives will be considered, and foreign direct investment (as well as other modes of foreign operation) will gradually take place in more distant countries.
One factor of overriding importance is the uncertainty perceived by decision-makers when entering foreign markets. Decisions about expansion into internationaloperations are for most firms, but particularly for those with limited international experience, characterized by a considerable amount of uncertainty. This uncertainty stems from two main sources. The first is a generallack of knowledge about the workings of particular foreign markets in terms of customer behavior, institutional framework and so on. The other is lack of knowledge of how to run a given business operation in an unfamiliar context. Inboth cases, the type of knowledge involved is typically acquired through a process of "learning by doing" (Carlson, 1975).
Another factor that influences this process is the availability of resources. Welch and Luostarinen (1988, p. 51) point out that "[T]he ability to undertake any form of international operations is clearly limited by the means accessible to the firm to carry it out. For smaller firms, given their limitations in many areas, this is an obvious reason why less demanding directions of international development can be undertaken first, with major commitments only occurring well into the longer run". The impact of company resources on internationalization has been documented in several studies. Studies offirms' export behavior indicate that firm size ispositively related not only to the propensity to export (Piercy, 1981), but also to the number of export markets served by an individual company (see for example Calof, 1994). Compared to exporting, foreign direct investments involve an even more substantial commitment of resources, both managerial and financial, to operations .that usually are considered as risky. Large firms should, due to their larger resource base, be in a better position than smaller firms to make such commitments, and it is not surprisingthat many studies report a positive relationship between the propensity to make foreign direct investments and the size of the firms (see for example Caves, 1974; Grubaugh, 1987).
Resources are also needed in order to absorb the costs and risk associated with FDI. For a given level of resources committed to an operation, the smaller the firm the more vulnerable it is ifsuch ventures turn out to be unsuccessful. Hence, small firms often take a cautious approach to international expansion (Welch and Luostarinen, 1988).
The process approach seeks to explain - and predict - two aspects of the internationalization of the firm. The first is the step-by-step fashion by which a firm's engagement in a specific country often develops. Although several stages are proposed in the literature, a typical establishment chain could begin with occasional exports, develop into regular exports via independent representatives (agent), followed by setting up sales subsidiaries, and end with fully-owned production facilities abroad. A study conducted by Newbould et al.(1978) on the internationalization of small British firms also showed that firms taking a cautious, stepwise approach generally performed better than firms that "leapfrogged stages". The worst performers were found to be firms that had skipped any intermediate stages and gone directly from not being involved in internationaloperations at all to establishing their own
20
production operations abroad. Due to the characteristics of the sample of firms studied (43 small and medium-sized companies) the results should of course be regarded as tentative.
Nevertheless, these findings underline the importance of knowledge and learning in the internationalization process. Without knowledge of and experience in a foreign market it is difficult to know how to operate .there, and the probability of failure increases accordingly.
The second aspect of the internationalization process is that firms are assumed to successively enter markets at an increasing distance from the home country, not only in terms of physical distance but also in terms of differences in economic development, language, culture, political system, etc. Thus, firms are predicted to start their internationalization by moving into markets they can most easily cope with, entering more distant countries only at a later stage.
Again, there are some indications that a cautious approach might payoff. For example, in a study of Dutch companies' internationalization, Barkema
et al.
(1993) report that prior experience in the same foreign country - or even in a fairly similar country (for example Norway and Sweden) - significantly increases the longevity of a foreign venture.To summarize this brief presentation of the currently dominant perspectives on the international behavior of firms; on one hand, the economics approach looks at companies' internationalization through the lens of a rational decision-making model, and contends that the various aspects of internationalization -like which markets are entered and how - can be regarded as rational, cost-minimizing responses to market imperfections and comparative cost factors. Onthe other hand, the behavioral approach views the internationalization of the firm from the perspective of organizational decision-making behavior characterized by uncertainty, limited search and experientiallearning, and describes the development of the various aspects of the internationalization of firms as one of "evolution" rather than
"revolution". These two approaches, which may seem to provide quite different - and perhaps even conflicting - perspectives on FDI behavior, constitute the theoretical basis for the analyzes conducted in the following studies of various aspects of companies' behavior regarding FDIs.
Objectives and Scope of the Studies
The overall objective of this dissertation is to advance the current knowledge about foreign direct investment and divestment behavior by looking at the FDI activities undertaken by Norwegian companies ..More specifically, the research conducted tests predictions about various dimensions of FDI behavior taken from both the economics and the internationalization process approaches to the internationalization of firms. The aim is to examine the explanatory value of current approaches, and thereby, if possible, clarify the relative strengths and weaknesses of the two approaches. Moreover, based on the empirical findings the research attempts to identify and suggest possible avenues for better conceptualizations of the issues under study.
One reason for choosing Norwegian FDI as the empirical setting for the studies is that apart from mainly descriptive reports, very few studies have been done on the foreign direct investment behavior of Norwegian companies. Among the exceptions are Karlsen and Randøy's (1991) empirical investigation - based on the "eclectic" theory of foreign direct investment - of determinants of FDI undertaken by large Norwegian companies, [uul and Walter's (1987) study - based on the internationalization process framework - of 12 Norwegian companies' FDI in the Ll.K. (however, only five of the FDls studied were made in manufacturing operations), and Smukkestad's (1979) study of Norwegian FDI in Southeast Asia. The scope of previous studies, and consequently the present state of knowledge about Norwegian foreign direct investment, is seemingly limited: they have looked at some isolated aspects of FDI like determinants of FDI, investigated the internationalization process leading to FDI for a very restricted number of companies, and/ or focused on FDls made in particular countries or regions.
However, although the current lack of any comprehensive studies of Norwegian FDI in manufacturing might by itself be considered as a sufficient reason for conducting the research, the main objective of the dissertation is not primarily to describe and analyze Norwegian FDI
per se,
but rather to make a contribution to the literature on the international22
operations of companies both theoretically and empirically. Norwegian data are well-suited in this respect. First, even though a number of studies of the FD! behavior of companies from small countries have emerged in recent years", the bulk of economic studies of FD! have focused on foreign direct investment in a North American context which - especially due to its characteristics with regards to size - is quite unique. Norwegian FD! provides an interesting empirical context because Norway is a small country with a limited domestic market, and most Norwegian multinational companies are, by international standards, quite small (to illustrate; only one Norwegian company is included in the overview of MNEs compiled by Stop ford and Dunning, 1983). Because of the small size of their domestic market, Norwegian companies have often had to expand internationally in order to achieve growth, but since the companies ingeneral are small they have had to do so with rather limited means. Thus, Norwegian FD! constitutes an empirical setting that in many respects deviate from the North American context. This gives an opportunity to assess whether explanations previously tested for U.S. FD! are valid in other contexts as well. Second, the Norwegian setting is also interesting from the viewpoint of the internationalization process framework which by-and-large originated from the findings of various studies of the internationalization of Swedish and Finnish firms. However, the predictions of the model with regard to foreign direct investments have not been tested on Norwegian data previously", Since the Nordic countries are fairly similar in many respects, it is often assumed that the internationalization process of the firms also is similar across the Nordic countries, i.e. that it follows the same general internationalization process model. Again, the Norwegian setting provides an opportunity to test the generalizability of the findings from previous studies.
There are four main limitations in the scope of the research. First, the empirical studies have been limited to deal with foreign direct investments in manufacturing, i.e. FD! relating to extractive, distribution or service activities are excluded. This has been done in order to ensure a certain degree of homogeneity in the unit of analysis. Moreover, establishment of foreign manufacturing operations are of particular interest because of the substantial commitment of resources involved in such FD!. The conventional definition of FD! as ownership of at least ten percent of the stock of a foreign company is applied throughout the
dissertation. Second, the scope of the research is limited in that it only covers investments done up to the mid-eighties. This limitation is largely due to availability of data. Third, the empirical studies focus exclusively on FDIs that have been made. They do not look into the actual FDI decision making process and many of the issues involved in this process like investment and acquisition negotiations, financing decisions, choice and adaptation of organizational structures, and personnel management. Finally, it should be noted that although the question of why and when foreign direct investment is chosen instead of some other form of foreign operation method (for example export) is discussed in some detail in one of the studies (chapter 5), the issue of FDI determinants is not analyzed empirically".
Research Questions and Framework for the Studies
The dissertation focuses on three particular aspects of foreign direct investments:
i) The location and expansion of foreign direct investments: the specific research question that will be examined is whether the location and expansion of foreign direct investments follow a gradual expansion path suggested by the internationalization process framework.
ii) The operation methods and ownership arrangements of foreign subsidiaries:
specifically, which factors determine the choice of ownership structure of Norwegian foreign affiliates in manufacturing? In addition, and more generally, to what extent are the current conceptualizations of the operation methods used by companies - including FDI - and explanations of their choice between the various methods, adequate?
iii) The divestment of foreign manufacturing operations: which factors determine whether the ownership involvement in a given foreign subsidiary is continued or dissolved over a period of time?
24
The research questions investigated inthis dissertation are "tied" together in that they taken as a whole constitute an - admittedly partial- "life cycle" view of foreign direct investments.
The overall structure of the dissertation is depicted in figure 1.1.
Figure .1.1. Framework for the studies.
External Factors inter alia
.. Distance
.. Economic Growth .. Political risk .. Development Level Internal Factors
inter alia .. Resources .. Experience .. Specific assets .. Relatedness
/
Internationalization of the Firm .. Location decisions
.. Operation method decisions .. Continuation decisions
to Entry t, Development tl Growth or retraction
The studies examine the location of the first FDIs undertaken, investigate the subsequent expansion routes taken by the companies with regard to the location of FDI, look at the choice of operation methods in general and the choice of ownership arrangements in particular, and finally, analyze the divestment (or conversely; "survival") of foreign manufacturing operations. As such the dissertation makes an effort toward dealing with the lack of longitudinal or dynamic approaches to foreign direct investments which has been noted by many (e.g. Horst, 1972; Larimo, 1993). More specifically, the longitudinal - or dynamic - dimension is taken into account in three different ways: i) chronological sequence of
expansions (chapter 2),ii) indusion of time (establishment year, age) as a variable or indicator in poly-item variable (experience) measurements (chapters 3,4 and 6), and iii) collection of data about the same units of analysis at different points in time (chapter 6).
Outline and Brief Summary of the Studies
The dissertation is composed of five individual studies. Two studies examine the location and expansion of foreign direct investments from the perspective of a gradual expansion model.
Then follows two studies that investigate and discuss the ways companies operate in foreign markets. One of the studies examines empirically the influence of various factors proposed by received theory on the choice of ownership structure of a foreign affiliate. The other study, which is of a conceptual character, provides an in-depth discussion of current research about foreign operation methods; and proposes some future lines of research. Finally, a study analyzes empirically why some foreign subsidiaries survive while others are divested. Inthe following, a short description of each study will be given.
The first study in this dissertation focuses on the location and expansion of foreign direct investments. The point of departure is that experience may affect the cost and the uncertainty of operating in foreign markets. Experience and market knowledge may therefore influence the location decisions of FDIs. Economic theory does not, however, predict a general expansion pattern of FDIs across industries. On the other hand, the theory associated with the internationalization process approach highlights the importance of cultural distance, and predicts a movement from "dose" to more "distant" locations as more experience is acquired by the firm. Two hypotheses are developed from the internationalization process approach regarding the locations of FDIs. First, it is proposed that the first FDIs undertaken by a company are made in countries that are culturally doser to the home country than later FDIs.
Second, it is proposed that the cultural distance to a country where a FDI is made will increase with the number of FDIs previously undertaken by a given company. The alternative hypothesis taken from the economics framework is that no general tendency to move into
26
distant countries should be expected as more experience is acquired. Instead, FDI location may be regarded as discrete decisions in which the "unfamiliarity" factor does not necessarily dominate other location factors. The hypotheses are tested on a data base consisting of a majority of Norwegian FDIs in manufacturing existing in 1982.Inall, 201 foreign direct investments are included in the data base. Cultural distance is measured by an index developed by Kogut and Singh (1988). The statistical results show no support for the notion that the first FDIs in general take place in culturally closer countries than later FDIs.
Moreover, for given companies, an expansion into more distant countries is not found as the number of investments increases. Thus, the internationalization process approach to location and expansion does not receive support.
The second study also focuses on the location of foreign direct investments, and hypotheses about the location of FDIs are again drawn from the internationalization process approach.
The central tenet of this framework is that location decisions should be regarded as a learning process at the company level. From this framework one would expect to find a close relationship between factors that increase the perceived level of uncertainty (such as distance), factors that serve to reduce uncertainty (such as experience), and factors that reduce the relative impact of the risk inherent in a project (the resources of the investing company), in the observed pattern of location choices. While the model tested in the first study was basically of a bivariate kind, the model tested in the second study is somewhat enlarged in that company resources is included as an additional predictor of firm behavior.
Moreover, additional concepts of distance - physical and economic distance - are introduced alongside cultural distance, and two different types of experience - general and specific experience - are taken into consideration. The hypotheses predict a positive relationship between the characteristics of the investing companies (resources and experience) and the distance (in economic, physical and cultural terms) to the chosen FDI locations. The hypotheses are tested on a data set consisting of 203 FDIs made by Norwegian companies in the period 1910 to 1984. The results provide limited support for the internationalization process framework. A positive relationship is found between the level of experience related to prior involvement in foreign manufacturing activities - termed specific experience in the
study - and distance to the chosen locations for FDI. Only weak support is found for a positive relationship between the export ratio of a company - measuring general experience - and distance. Furthermore, no support is found for the hypothesized positive relationship between company resources and distance to where a foreign manufacturing subsidiary is located. The results are very similar across regressions with different specifications of the dependent variable (economic, physical, and cultural distance). However, the clearest results were obtained for physical distance as the dependent variable. Overall, the results suggest that the internationalization process model is a rather partial model, and that it needs to be supplemented by economic and strategic variables in order to explain the location of foreign direct investments.
The third study in the dissertation looks at how multinational enterprises establish foreign .subsidiaries. Previous studies on foreign direct investment and multinational enterprises
have mainly focused on why companies choose to establish foreign production subsidiaries rather than exploiting their firm-specific advantages by exporting. However, once a company has decided to invest abroad by establishing a manufacturing unit in a foreign country it must also choose an appropriate ownership structure of the subsidiary. The two main alternatives are either a wholly-owned foreign affiliate or a joint venture with another partner. The question of ownership has important ramifications both in terms of the level of control a company has over a foreign operation and the flexibility it has to reallocate the assets if necessary. Hypotheses regarding the choice between wholly-owned and partly- owned subsidiaries are drawn from both economic (transaction cost theory) and behavioral (internationalization process) perspectives. The hypotheses propose that the propensity to choose a wholly-owned subsidiary will increase, i) the larger the resource base of the firms, ii) the more experienced firms are, and iii) the higher the importance of proprietary assets.
On the other hand, the propensity to wholly-own foreign manufacturing subsidiaries is expected to decrease, i)the larger the cultural distance to a host country, and ii)the higher the political risk of a host country. The hypotheses are tested on a sample of 174 foreign direct investments made by Norwegian companies, and still owned by these companies in 1984.
One main finding is that the political risk associated with the host country strongly increases
28
the probability that ownership of a foreign subsidiary is shared. This result suggests that under risky circumstances companies are willing to trade-off the benefits of control for a higher degree of strategic flexibility. Inaddition, taking a local partner into a joint venture may also reduce political complications and the risk of being expropriated. Another finding is that large cultural distance between the home and the host countries leads to a higher propensity to enter into joint venture arrangements. This result supports the notion that a large cultural distance increases the uncertainty perceived by decision-makers, and makes it more difficult for an entrant to know how to run an operation successfully. Inorder to overcome the unfamiliarity with market conditions, cultural traits, etc., knowledge about local conditions is needed which in tum can be made accessible by teaming-up with a local firm. However, as firms get more experience from foreign operations one might expect that they would become less dependent on other firms as providers of the necessary knowledge.
The results provide some support for this line of reasoning for investments of a vertical kind.
Onthe other hand, the ownership structure of horizontally related foreign subsidiaries is not influenced by the Norwegian parent companies' level of international experience. Finally, little support is found for a transaction cost approach to the choice of ownership arrangement. The coefficients of the proxies for proprietary content are insignificant (and even in the direction opposite to the one expected) in a majority of the regressions. Overall, the main insight from the study is that the conduct of Norwegians companies appears to be largely determined by external factors, in particular the political risk of the host country.
The fourth study presents a further elaboration on the issue of how companies enter and operate in foreign markets. An overview and critique of the two main approaches to this issue - the economics approach and the so-called internationalization process approach - is undertaken. While considerable progress has been made in both streams of literature, substantial deficiencies still exist. The economics approach is criticized for being a rather restricted and simplifying framework with regard to organizational decision-making behavior and the degree of rationality assumed to characterize decision-makers. Moreover, contributions based on this framework typically treat any given operation method as characterizable in terms of specific and objective levels of control, risk, resource
commitments, etc., which again provide the information required for classification. However, the real-life complexity of actual operation modes often makes it difficult to classify them accurately. Besides, the perceived levels (which even from an economic perspective should be decisive) of control, risk, etc., offered by a given operation method, may in fact vary considerably across different companies. Finally, the economics approach is rather static, focusing foremost on how rational economic actors (are assumed to) choose a more or less
"optimal" mode of entry into a given market at a given point in time. Little attention is paid to decisions about changes to the initial entry modes, and to how such decisions interact with other aspects of the internationalization of the firm. The alternative approach - the internationalization process framework - places great emphasis on behavioral factors like experience, knowledge, and perceived risk, as driving forces in the internationalization of firms in general and their use of various operation methods in particular. The framework depicts the "choice of operation mode" as one of gradual development, i.e. a move from low- commitment to high-commitment modes over time, often described as an "establishment chain". Although the internationalization process framework represents a more micro- analytic and process oriented approach to the study of the behavior of firms than the
"economics" framework, it does not escape criticism. First, while the economics approach may have little to say about dynamics, the internationalization process framework can - especially in the early contributions - be criticized for describing the development of firm's internationalization in overly deterministic terms. Thus, longitudinal processes (Le.
dynamics) are certainly focused upon, but the general implication of the analysis seems at the same time to have been one of an inescapable incremental path (i.e. determinism). Close attention to the processes at work would most likely show a considerable diversity in the operation methods used by different companies. Anincreasing number of empirical studies give, not surprisingly, support to the view that the internationalization of firms cannot, in general, be described as one of gradually increasing commitments: "leapfrogging" as well as
"reversals" appear to be quite common. Second, the internationalization process approach has paid little attention to how factors beyond those closely linked to the organizational decision- making process
per se
may influence the outcome of such processes. However, many other influences - both internal and external to the firm - may be operative. For example, internal30
situational factors like the financial state of a company or the current utilization level of production capacity are likely to shape any decision taken, and so will external factors such as the competitive situation in an industry or market. Finally, the internationalization process framework has, like the "economics" framework, paid little attention to the increasing complexity of the operation methods actually in use. Frequently, companies do not face a choice between (a limited number of) different operation methods. The challenge is rather to put together an appropriate package of methods in order to operate in a particular foreign location. Taken together, the preceding concerns with the existing approaches to the choice of foreign operation mode suggest that a considerable research effort lies ahead. At the conceptuallevel, a much better understanding of the mechanisms driving both the initial entry modes and the subsequent changes of operation mode packages is needed. Attention should be given to both the internal and the external context in which such decisions are made. Given the complexity and longitudinal nature of these phenomena, qualitative methodologies seem particularly appropriate for future research.
While the first four studies investigate the growth and expansion of companies in international markets, the last study in the dissertation takes a look at the other side of the coin, that is; to what extent and why are foreign units divested? Foreign direct investment represents, in principle, a long-term commitment to a foreign operation. Divestments appear nevertheless to be quite common. For example, Barkema et al.(1993) conducted a study of the longevity of foreign direct investments made by the largest Dutch multinational enterprises.
They report that of 225 FDls made in the period 1966 to 1988, only half of these were still in existence in 1988. However, besides the study by Barkema et al.(1993) and a few other studies (e.g. Shapiro, 1986) the question of what might influence whether foreign subsidiaries are divested or not is largely unexplored. The study investigates some determinants of Norwegian companies' divestment of foreign manufacturing operations. The perspective taken in the study is that divestments can be regarded as a function of two factors: Incentives to exit from an operation and barriers to exit. The model includes factors that might lower or heighten mobility barriers - and hence increase or decrease the propensity to divest - suggested by industrial organization theory, strategic management literature, and the
behavioral approach to the internationalization of the firm. The study is designed as a ten- year follow-up study with observations taken at two points in time; first in 1982 and then in 1992. Foreign units are considered as divested ifthey were no longer owned by the same Norwegian parent company at the end of the period. The study shows that more than half of the foreign subsidiaries existing in 1982 were divested within a period of ten years. Among the factors examined in the study, three factors turned out to be of particular importance for the decision to retain or divest foreign units. First, economic growth in the host country increase the probability that operations will be continued. Second, foreign entries by acquisition face a much higher risk than greenfield ventures for subsequent dissolution.
Finally, the probability of foreign divestment increases with the size of the parent company.
After the five studies, a final chapter follows where the main findings are discussed. That chapter also contains a discussion of the contributions of the studies and of the limitations of the research. Some suggestions for future research closes the dissertation.
